Sampling procedures for estimating within-tree populations of second-stage larvae (L2) of Matsucoccus feytaudi Duc., infesting maritime pine (Pinus pinaster Ait.), were investigated. These included random sampling without replacement, and systematic sampling with and without a linear model. The relative precision of the sampling was affected by the number, size, and bark thickness of the sample units. Because of symmetrical vertical distribution of within-tree populations, systematic sampling with a linear model did not increase precision when compared with simple systematic sampling. Sampling can be profitably reduced to below the crown portion of the bole, after removal of the part with either smooth or very thick and pyramidal bark. Counting L2 exuviae in the upper 10 cm of each 20-cm-long log, using one in every three logs, provided a relative precision of about 40%. The numbers of male M. feytaudi caught in sticky traps baited with 5 or 30 μg of synthetic pheromone were compared with the numbers of L2 estimated according to the sampling method previously developed. There was a significant positive correlation between number of scales caught and L2 estimates, at the level of individual trees for the lower dose lures, and at the level of groups of trees for the higher dose lures, used in 20-year-old stands. Large captures in younger, weakly infested stands were related to a possible immigration of flying scale insects.